Bug 46275 - KeepAliveTimeout in milliseconds
Summary: KeepAliveTimeout in milliseconds
Status: RESOLVED FIXED
Alias: None
Product: Apache httpd-2
Classification: Unclassified
Component: Core (show other bugs)
Version: 2.5-HEAD
Hardware: All All
: P4 enhancement (vote)
Target Milestone: ---
Assignee: Apache HTTPD Bugs Mailing List
URL:
Keywords: PatchAvailable
Depends on:
Blocks:
 
Reported: 2008-11-24 05:25 UTC by Takashi Sato
Modified: 2009-01-11 16:55 UTC (History)
0 users


Attachments
against trunk (1.78 KB, patch)
2008-11-24 05:25 UTC, Takashi Sato 		Details | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Takashi Sato 2008-11-24 05:25:44 UTC 
Created attachment 22921 [details]
against trunk

While keepalive can save signifigant overhead in creating TCP connections it can waste workers. So KeepAliveTimeout is important.
I believe 1 seconds of KeepAliveTimeout is long because most file can be sent in under 1 second.
Comment 1 Eric Covener 2008-11-24 07:15:47 UTC 
> I believe 1 seconds of KeepAliveTimeout is long because most file can be sent
> in under 1 second.

Isn't the KeepAlive timout not ticking away while a file is actually being transferred?
Comment 2 Takashi Sato 2008-11-24 07:29:44 UTC 
(In reply to comment #1)
> > I believe 1 seconds of KeepAliveTimeout is long because most file can be sent
> > in under 1 second.
> 
> Isn't the KeepAlive timout not ticking away while a file is actually being
> transferred?
> 

No, it isn't.

What I want to say is the comparison of KeepAliveTimeout and file serving.
While one worker waits one second, other worker serves several files. I think this is waste of workers.
Comment 3 Takashi Sato 2009-01-11 16:55:52 UTC 
Comitted as r733557